Spatial Coherence Beamforming in Lung Ultrasound Imaging from Numerical Simulations to Ex Vivo Large Animal Data

Lung ultrasound imaging plays a crucial role in Point-of-Care diagnostics of pulmonary lesions. Modern clinical imaging is aimed for quantitative, reproducible, real-time, radiation-free and portable methods for general and localized assessment of lung aeration. Reproducibility of hand-held transthoracic lung ultrasound scanning along with agreement of human readers on B-mode images scoring are limited. Often developed methods encounter obstacles in translation from numerically simulated to ex vivo and in vivo data. Two RF datasets comprised of 1000 numerically simulated scans and 120 ex vivo porcine tissue scans of known ground truth (lung aeration) were obtained. We studied brigntness distributions in beamformed images comparing performance of conventional Delay-and-Sum (DAS) and developed Short Lag Spatial Coherence (SLCS) beamforming techniques. Translatability of developed lung-specific SLSC beamformer from numerically simulated to ex vivo data is demonstrated here.